Image processing

Abstract

In an image-processing method, a stack is provided for storing a predetermined number of frame portions. An image including a target object is obtained, the image being formed by an array of pixels. A frame portion is extracted from the image, the frame portion being at least a portion of the pixels forming the image, corresponding to a region of interest in the image, the region of interest comprising the target object. The frame portion is stored in the stack, the storing including discarding an oldest previously stored frame portion from the stack if the number of frame portions stored in the stack has reached the predetermined number. The steps of the method are repeated a plurality of times. Frame portions in the stack having a phase substantially equal to a given phase are averaged. A super-resolved image is calculated from the plurality of stored frame portions.

Description

FIELD OF THE INVENTION[0001]This invention relates to the field of image processing, and in particular to the generation of a higher-resolution image from a plurality of lower-resolution images. The invention relates especially, but not exclusively, to the field of image-processing in homing missiles.BACKGROUND OF THE INVENTION[0002]The accuracy of detection, recognition, identification and tracking using electro-optical sensors is critically dependent upon image quality; hence improved image quality can be expected to result in increased performance. However, in some applications there is a need to strike a balance between, on the one hand, obtaining the best possible image quality and, on the other hand, obtaining an image quickly and with processor and other hardware requirements that are not too demanding. For example, in the field of homing missiles, provision of images of higher quality than those provided by existing systems would increase the distance over which an image pro...

AI Technical Summary

Benefits of technology

[0022]It may be that the calculation of the super-resolved image includes enlarging the averaged regions of interest by a super resolution scale factor and averaging pixels across all phases, accounting for subpixel shifts. It may be that calculating a super-resolved image from the plurality of stored frame portions includes updating a super-resolved image calculated in a previous iteration by changing in the calculation only the phases which have changed in a new frame. It may be that calculating a super-resolved image from the plurality of stored frame portions includes updating a super-resolved image calculated in a previous iteration by removing the oldest frame from its corresponding phase average, adding the new frame to its corresponding phase average, and updating the phase average over phases with the two modified phases. It will be appreciated that by removing the oldest frame, having a first phase, and adding a new frame having a second (typically different) phase, two of the phase averages will be calculated using a different number of frame portions as compared to the previous calculation. As such, the effective weighting given to the individual frame portions as combined to create the super-resolved image will typically vary. It may be that a translation shift between a first pair of frame portions is calculated and the resulting calculated translation shift is used in calculating a first super-resolved image derived from a first set of frame portions including said first pair of frame portions. Subsequently, a second super-resolved image may be calculated from a second set of frame portions, different from the first set, but still including said first pair of frame portions, wherein the second super-resolved image is calculated using the previously calculated translation shift. There may then be no need to recalculate the previously calculated translation shift when calculating the second super-resolved image, resulting in more efficient image processing. The previously calculated translation shift may originally be calculated in a manner other than comparing a pair of frame portions. For example, the relative registration of a pair of successive frame portions may be computed from a measure of tracking. Not having to recalculate a previously calculated translation shift may be of advantage in embodiments of the invention of a type for which the traditional first step of super-resolution processing would be to compute the registration of all the images to a single master image.

Problems solved by technology

However, known super-resolution techniques, for example in astronomy, are slow and require a great deal of computer processing power.

Images